基于蚁群-模糊聚类算法的井下工作面瓦斯突出预测

Gas Outburst Prediction of Underground Working Face Based on ACA-FCM Algorithm

针对现有的瓦斯预测方法在实际应用中受到较大限制且预测结果的准确性较差的问题,提出了一种基于蚁群-模糊聚类算法的瓦斯突出预测方法;分析了蚁群-模糊聚类算法的基本原理及实现步骤,并以某煤矿井下工作面某时段内的瓦斯突出数据为例,采用蚁群-模糊聚类算法对该数据进行了挖掘分析,从而找出了瓦斯突出量与其影响因素即埋藏深度、煤层厚度、瓦斯含量、日进度、煤层间距、日产量之间的关系。测试结果表明,该方法预测结果与实际监测记录完全一致,具有较高的聚类预测性能。

In view of problems of great limitation in actual application and bad precision of predictionresult of current gas outburst prediction method, the paper proposed a gas outburst prediction method based on ACA-FAM algorithm. It analyzed basic principle and implementation steps of ACA-FAM algorithm. Taking data of gas outburst of underground working face of a Coal Mine in a certain period as example, it used ACA-FAM algorithm to make mining analysis for the data to find relations between gas outburst and influencing factors such as buried depth, coal seam thickness, gas content, daily advance, coal seam interval and daily output. The test result shows that prediction result of the method is uniform with actual monitoring record and the method has higher classified prediction performance.